A School With No Teachers Costs $55,000.

Alpha Schools is enrolling students in Chicago for fall 2026. The model is K through 8. The tuition is $55,000 per year. And there are no teachers.

Instead, each student gets a "guide," an adult who facilitates but does not instruct. Students spend one to two hours each day on a computer, learning core academic subjects entirely through AI programs. The rest of the day is filled with guide-led workshops in public speaking, coding, outdoor education, and other enrichment activities.

CBS News covered the expansion this week. The reporting is measured, and it should be. This is not a gimmick. Alpha Schools is sincere. They believe, with real conviction, that AI can deliver personalized academic content more effectively than a single teacher managing twenty-five students at once. And in a narrow sense, they may be right. An AI tutor can adapt pacing, repeat explanations without fatigue, and meet each student exactly where they are on any given concept.

But "delivering content" and "teaching" are not the same thing. And the distance between them is where everything important lives.

What a Guide Is Not

A guide facilitates. A teacher does something harder to name.

A teacher notices when a student's confidence shifts mid-paragraph. A teacher asks a question not because the rubric demands it, but because they sense a student is close to something they have not articulated yet. A teacher holds the tension between what a student knows and what they are about to discover, and they hold it long enough for the student to cross that gap on their own.

This is not sentimentality. It is pedagogy.

The OECD's Digital Education Outlook 2026 showed that general-purpose AI boosts student performance by 48%, but creates a 17% deficit when the AI is removed. The students were carried, not coached. They performed better with the tool and worse without it because the tool had been doing the thinking for them, so seamlessly they never noticed it happening.

The relationship between a student and a well-designed question, asked by someone or something that understands where the student is and where they need to go, is the engine of real learning. Remove that relationship and you can still deliver content. You can still produce test scores. But you lose the thing that makes education transformative.

The $55,000 Question

The price tag reveals something important. This is not a model designed to democratize education. It is a premium product for families who can afford it.

And it raises a question about equity that is difficult to set aside. If AI-only education costs more than most private schools in America, what happens to the 90% of students in public schools? Do they get the cheaper version? A stripped-down AI tutor with no guide, no outdoor education, no enrichment workshops?

The answer cannot be "they get less." The answer has to be that we build tools that make great teaching more visible and more scalable. Not tools that replace teachers with software and call it innovation. Not tools that work beautifully for the families who can write a $55,000 check and leave everyone else behind.

What AI Should Actually Do in a Classroom

The best use of AI in education is not to replace the teacher. It is to extend what the teacher can see.

Right now, a teacher with 120 students has no way to watch each one think. The process between assignment and submission is invisible. A student might spend three hours wrestling with an argument, revising it twice, abandoning a weak thesis, and arriving at something genuinely original. Another student might paste a prompt into a chatbot and submit whatever comes back. In most classrooms, those two submissions look identical.

AI can change that. Not by doing the teaching, but by making the learning visible.

This is the approach we have built at Koan. Our AI tutor, Aidan, does not teach students. It asks them questions. Socratic questions calibrated to their rubric and adapted to their patterns of thinking. And every revision, every pause, every shift in reasoning is captured so that teachers can see what happened between the blank page and the final draft. The teacher is not replaced. The teacher is empowered with evidence they never had before.

A Reflection

Alpha Schools is asking a version of the right question: how do we personalize learning at scale? That question deserves serious engagement, not dismissal. But they have arrived at an answer that removes the very thing that makes learning transformative: the human relationship between a student and someone who cares whether they understand.

The future of education is not teacherless classrooms. It is not AI-free classrooms either. It is classrooms where technology makes visible what was always hidden, so that teachers can do what only teachers can do: notice, respond, and hold space for the kind of thinking that changes a person.

A school without teachers can still produce test scores. It can still move students through a curriculum. But it cannot do the thing that a great teacher does in a single well-timed question, the thing no algorithm has yet replicated: make a student feel seen in the exact moment they are becoming someone new.

If a school can function without teachers, what does that tell us about what we were asking teachers to do in the first place?